5524
Langmuir 2005, 21, 5524-5528
Protective Effects of Small Amounts of Bis(2-ethylhexyl)sulfosuccinate on the Helical Structures of Human and Bovine Serum Albumins in Their Thermal Denaturations Yoshiko Moriyama and Kunio Takeda* Department of Applied Chemistry, Okayama University of Science, 1-1 Ridai-cho, Okayama 700-0005, Japan Received January 28, 2005. In Final Form: April 11, 2005 The protective effect of an anionic double-tailed surfactant, sodium bis(2-ethylhexyl)sulfosuccinate (AOT), on the structures of human serum albumin (HSA) and bovine serum albumin (BSA) in their thermal denaturations was examined by means of circular dichroism measurements. The structural changes of these albumins were reversible in the thermal denaturation below 50 °C, but became partially irreversible above this temperature. The effect was observed in the thermal denaturation above 50 °C. Although the helicity of HSA decreased from 66% to 44% at 65 °C in the absence of the surfactant, the decrement of it was restrained in the coexistence of AOT of extremely low concentrations. When the HSA concentration was 10 µM, the maximal protective effect appeared at 0.15 mM AOT. In the coexistence of the surfactant of this concentration, the helicity was maintained at 58% at 65 °C, increasing to the original value upon cooling to 25 °C. Beyond 0.15 mM AOT, the helicity sharply decreased until 3 mM AOT. A particular AOT concentration required to induce the maximal protective effect ([AOT]REQ) was examined at different HSA concentrations. [AOT]REQ shifted to higher values with an increase of the protein concentration. From the protein concentration dependences of [AOT]REQ, the maximal protection was estimated to require 8.0 and 5.0 AOT ions per a molecule of HSA and BSA, respectively. The AOT concentration, where the protective effect was observed, was too low to form its micelle-like aggregate. Then the protein structures might be stabilized by a cross-linking of surfactant monomers bound to specific sites. These specific sites might exist between a group of nonpolar residues and a positively charged residue located on several sets of amphiphilic helical rods in the proteins. Such a unique function of the double-tailed ionic surfactant is first presented by its characteristic nature as an amphiphilic material.
Introduction Many studies have extensively been carried out on the interactions between surfactants and proteins for half a century.1-5 However, little attention has been paid to a certain protective effect of ionic surfactants such as sodium dodecyl sulfate (SDS) on protein structure.6-11 This typical effect of SDS appears upon the urea denaturation of serum albumins: the coexistence of a small amount of SDS protects the helical structures against the urea denaturation,6-10 or the helical structures, lost in the urea denaturation, can be almost refolded9,10 upon the addition of such an amount of the surfactant. The recovered helicity * To whom correspondence should be addressed. E-mail: takeda@ dac.ous.ac.jp. Phone and fax: 086-256-9553. (1) Steinhardt, J.; Reynolds, J. A. Multiple Equilibria in Proteins; Academic Press: New York, 1969; pp 239-302. (2) Jones, M. N. Biological Interfaces; Elsevier: Amsterdam, 1975; pp 101-130. (3) Lapanje, S. Physicochemical Aspects of Protein Denaturation; Wiley-Interscience: New York, 1978; pp 156-179. (4) Takeda, K.; Moriyama, Y.; Hachiya, K. In Encyclopedia of Surfactant and Colloid Science; Hubbard, A. T., Ed.; Marcel Dekker: New York, 2002; Part 1, pp 2558-2574. (5) Takeda, K.; Hachiya, K.; Moriyama, Y. In Encyclopedia of Surfactant and Colloid Science; Hubbard, A. T., Ed.; Marcel Dekker: New York, 2002; Part 2, pp 2575-2592. (6) Duggan, E. L.; Luck, F. M. J. Biol. Chem. 1948, 172, 205-220. (7) Markus, G.; Karush, F. J. Am. Chem. Soc. 1957, 79, 3264-3269. (8) Markus, G.; Love, R. L.; Wissler, F. C. J. Biol. Chem. 1964, 239, 3687-3693. (9) Moriyama, Y.; Sato, Y.; Takeda, K. J. Colloid Interface Sci. 1993, 117, 420-424. (10) Moriyama, Y.; Takeda, K. Langmuir 1999, 15, 2003-2008. (11) Moriyama, Y.; Kawasaka, Y.; Takeda, K. J. Colloid Interface Sci. 2003, 257, 41-46.
of bovine serum albumin (BSA) is 62%9 as compared to 66% in the native state,4,9,11,12 50% at high SDS concentrations,4,9,11,12 and 10-20% at high urea concentrations.9,13 (The term “helicity” will be used throughout this paper as a shorthand notation for the cumbersome phrase “extent of R-helical structure”.) In the case of human serum albumin (HSA), the helicity is recovered or protected up to the original level.10 The protective phenomenon of SDS was first found by Duggan and Luck by observing that the addition of the surfactant prevents the rise in viscosity of serum albumins in urea solutions.6 Although refolding processes of proteins have recently been studied, little is known concerning the refolding of a protein induced by the function of a surfactant. Except the aforementioned studies, only Markus and Karush7 and Markus et al.8 studied the protective effect of surfactants on the urea denaturation of HSA by means of optical rotatory dispersion methods. Recently, we reported that the protective effect of a small amount of SDS also appears on structures of BSA in the thermal denaturation.11 On the other hand, several surfactants have been used as tools to isolate, solubilize, and manipulate many proteins for subsequent biophysical and biochemical characterization.14,15 These methods skillfully depend on the unique nature of the surfactant as an amphiphilic material. In these processes, however, the role and (12) Takeda, K.; Shigeta, M.; Aoki, K. J. Colloid Interface Sci. 1987, 117, 120-126. (13) Takeda, K.; Sasa, K.; Kawamoto, K.; Wada, A.; Aoki, K. J. Colloid Interface Sci. 1988, 124, 284-289. (14) Waehneldt, T. V. BioSystems 1975, 6, 176-187. (15) Garavito, R. M.; Ferguson-Miller, S. J. Biol. Chem. 2001, 276, 32403-32406.
10.1021/la050252j CCC: $30.25 © 2005 American Chemical Society Published on Web 05/13/2005
SDS Protects BSA Structure in Heat Denaturation
mechanism, which the surfactants play as tools, are not well defined yet, despite many studies of proteinsurfactant interactions. A further study of the interaction is necessary not only to deepen the understanding of the interaction but also to know the mechanism by which the amphiphilic materials, such as a lipid, work on proteins in a biological system. In the present study, we have mainly examined the protective effect of sodium bis(2-ethylhexyl)sulfosuccinate (AOT) on the structure of HSA in the thermal denaturation. The thermal denaturations of the serum albumins have a characteristic aspect: the conformational changes are reversible below 50 °C, but are only partially reversible above this temperature.11,16-18 The present study shows that these protein structures can be almost protected from the irreversible conformational changes in their thermal denaturations upon the coexistence of small amounts of AOT. It becomes clear that this protective effect is induced by the anionic surfactant in a monomer state,11 although a surfactant is generally recognized to fulfill its function in the aggregated state such as a micelle, membrane, layer, and so on or to exhibit a characteristic function by forming a micelle-like aggregate on a protein or polymer. The present surfactant has the anionic hydrophilic group and two hydrophobic chains. This double-tailed surfactant structurally resembles phospholipids, which are naturally occurring surfactants. Then the present study might also give a hint in the consideration of a new role of such lipids in addition to the formation of bilayers or vesicles in vivo. Materials and Methods HSA and BSA were purchased from Sigma Chemical Co. (A378210 and A1900,9 respectively). AOT19 and SDS9-11 were obtained from Aldrich Co. and Fluka Chemie AG, respectively. The weighing of AOT powder was carried out in an atmosphere of nitrogen.19 A sodium phosphate buffer of pH 7.0 and ionic strength 0.01411,12,20 was exclusively used to prepare each solution. The concentrations of HSA and BSA were determined spectrophotometrically using 280 ) 35000 and 44000, respectively.21 The circular dichroism (CD) measurements were carried out with a Jasco J-720 spectropolarimeter at various temperatures up to 65 °C. The protein solution was put into a 1.0 mm path length cell with a water jacket, through which water was circulated at a desired temperature. The CD spectrum was measured after the protein solution was kept at the desired temperature for 30 min.11 When the temperature of the circulating water was raised or cooled to another temperature through the measurements or when the protein solution was kept at a certain temperature, the cell containing the protein solution was protected from the ultraviolet beam. This is because the irradiation of such an ultraviolet light disrupts the structure of protein.22,23 The helicity was estimated by the curve-fitting method of the CD spectrum, using the reference spectra as determined by Chen et al.24 The simulation was carried out in the wavelength region 200-240 nm at 1 nm intervals. The CD spectra of serum albumins can be well simulated4,9-13,19,20,25,26 by using the reference spectra of Chen et al. In the present study, (16) Imahori, K. Biochim. Biophys. Acta 1960, 37, 336-341. (17) Anderle, G.; Mendelsohn, R. Biophys. J. 1987, 52, 69-74. (18) Lin, V. J. C.; Koenig, J. L. Biopolymers 1976, 15, 203-218. (19) Takeda, K.; Harada, K.; Yamaguchi, K.; Moriyama, Y. J. Colloid Interface Sci. 1994, 164, 382-386. (20) Takeda, K.; Sasa, K.; Nagao, M.; Batra, P. P. Biochim. Biophys. Acta 1988, 957, 340-344. (21) Sober, H. A., Harte, R. A., Eds. Handbook of Biochemistry (Selected Data for Molecular Biology), 2nd ed.; CRC Press: Cleveland, OH, 1973; p C-71. (22) Takeda, K.; Moriyama, Y. J. Am. Chem. Soc. 1991, 113, 67006701. (23) Moriyama, Y.; Kakehi, T.; Takeda, K. Anal. Biochem. 1994, 219, 378-380. (24) Chen, Y. H.; Yang, J. T.; Chau, K. H. Biochemistry 1974, 13, 3350-3359.
Langmuir, Vol. 21, No. 12, 2005 5525 the helicity is discussed using the values calculated by the curvefitting as a measure of helicity,9-13 although the helicity has frequently been discussed on the basis of the change of [θ]222.
Results and Discussion The R-helicities of both HSA10 and BSA4,9-13,19,20,25,26 have been estimated to be 66% from the CD spectra. The homologous proteins HSA and BSA are single polypeptides consisting of 58527 and 58328 amino acid residues, respectively. The helicity of 66% is reasonable compared with the values determined by other groups29-31 and the helicity of 67% determined for HSA by the recent X-ray crystallographic study.32,33 The helicity of 66% indicates that the helical structures are formed at about 385 amino acid residues in the total residues of these protein molecules. On the other hand, it is generally understood that the structural changes of HSA and BSA are reversible in the thermal denaturation below 50 °C, but become partially reversible above this temperature.11,16-18 Recently, we examined the thermal denaturation of BSA by the CD method; the helicity of 66% at 25 °C decreases to 44% at 65 °C, and it does not attain the original value upon cooling from 65 to 25 °C, increasing only to 53%.11 The thermal denaturation of HSA was also examined in the present study. As a result, the same profile of the secondary structural change was observed as a function of temperature between 25 and 45 °C and between 25 and 65 °C. The difference between the 66% helicity at 25 °C and the 53% helicity at 65 °C corresponds to the helices formed at 76 amino acid residues. This means that about 20% (76/385 originally helical residues) of the original helices in these proteins are not recovered upon the descent of temperature. As compared with this, the helicity decreases to 61% with a rise of temperature up to 45 °C. This helicity at 45 °C completely recovers to the original value upon cooling to 25 °C. 11 Figure 1 shows the dependence of the helicity of HSA on the AOT concentration at 25 °C. The helicity of HSA sharply decreased until 2 mM AOT. The CD spectra of HSA were measured under the coexistence of AOT of various concentrations at several temperatures to examine the AOT effect on the protein structure in the thermal denaturation. Figure 2 shows the CD spectra of HSA at 65 °C under the coexistence of AOT of several concentrations. The presence of 0.15 mM AOT mostly restored the CD spectrum indicative of the disordered form to the spectrum indicative of R-helical structure. However, a higher concentration of the surfactant caused the distortion of this R-helical-type spectrum. Figure 3 shows the dependence of the helicity of HSA on the AOT concentration at 65 °C. Although the helicity decreased from 66% to 44% at 65 °C in the absence of the surfactant, the decrement of it was restrained at lower AOT concentrations, as seen in Figure 3. Beyond 0.15 mM AOT, the helicity gradually and sharply decreased until 2 mM AOT. (25) Takeda, K.; Wada, A.; Nishimura, T.; Ueki, T.; Aoki, K. J. Colloid Interface Sci. 1989, 133, 497-504. (26) Batra, P. P.; Sasa, K.; Ueki, T.; Takeda, K. Int. J. Biochem. 1989, 8, 857-862. (27) Brown, J. R. In Albumin Structure, Function, and Uses; Rosenoer, V. M., Oratz, M., Rothschild, M. A., Eds.; Pergamon Press: Oxford, 1977; pp 27-51. (28) Hirayama, K.; Akashi, S.; Furuya, M.; Fukuhara, K. Biochem. Biophys. Res. Commun. 1990, 173, 639-646. (29) Reed, R. G.; Feldhoff, R. C.; Clute, O. L.; Peters, T., Jr. Biochemistry 1975, 14, 4578-4583. (30) Geisow, M. J.; Beaven, G. H. Biochem. J. 1977, 163, 477-484. (31) Wetzel, R.; Becker, M.; Behlke, J.; Billwitz, H.; Bohm, S.; Ebert, B.; Hamann, H.; Krumbiegel, J.; Lassmann, G. Eur. J. Biochem. 1980, 104, 469-478. (32) He, X. M.; Carter, D. C. Nature 1992, 358, 209-215. (33) Carter, D. C.; Ho, J. X. Adv. Protein Chem. 1994, 45, 153-203.
5526
Langmuir, Vol. 21, No. 12, 2005
Figure 1. Helicity change of HSA in the AOT soluition at 25 °C.
Figure 2. CD spectra of HSA at 65 °C upon the coexistence of AOT of several concentrations. The numerical values in the figure denote the added AOT concentrations (mM). The broken line indicates the spectrum of the native HSA at 25 °C without AOT.
Figure 3. Dependence of the helicity of HSA on the AOT concentration in the thermal denaturation at 65 °C.
The protective effect appears only below 2 mM AOT. In other words, the effect of the thermal denaturation on the helicity is more restrained below 2 mM AOT than above this concentration. It should be noted that the great protective effect appears in the coexistence of AOT of extremely low concentrations. The expanded profile of the helicity change at low AOT concentrations at 65 °C is shown in Figure 4. This figure gives a particular AOT concentration required to induce the maximal protective effect ([AOT]REQ). The protective effect grew strong with an increase in AOT concentration in its low concentration region, and [AOT]REQ was about 0.15 mM. The protein
Moriyama and Takeda
Figure 4. Helicity change of HSA at low AOT concentrations at 65 °C (O). Closed circles designate the helicities upon cooling to 25 °C after the thermal denaturation at 65 °C.
Figure 5. Dependence of the helicity change of HSA as a function of AOT concentration on the protein concentration (0, 20 µM, 4, 30 µM, ], 40 µM) in the thermal denaturation at 65 °C. The broken line indicates the helicity change at 10 µM HSA.
concentration was 0.01 mM, and then the molar ratio of [AOT] to [HSA] of about 15 was enough to induce the maximal protective effect on the helicity of HSA. In addition, the helicity of HSA was completely recovered to the original value at this molar ratio upon cooling to 25 °C after the heat treatment, as is also shown in Figure 4. The maximal protected helicity was 58% at 65 °C. This helicity increased to 66% upon cooling to 25 °C. Indeed, most of the original helices are protected by such a small amount of AOT. The protected helicity contains most of the aforementioned 20% of the original helices (corresponding to the helices formed at 76 residues), which are not reversibly re-formed after the heat treatment at 65 °C. This fact should be emphasized, because the helicity of BSA was not recovered to the original value but to the 62% level upon cooling.11 Such a difference in stabilities of the helicities of these two proteins has been observed also in the urea denaturation.9,10 The protective effect of AOT was examined at different HSA and BSA concentrations. The concentration of [AOT]REQ increased with an increase of HSA concentration, as shown in Figure 5. Figure 6 shows the changes of [AOT]REQ as functions of the concentrations of HSA and BSA. Linear relations were substantially obtained between [AOT]REQ and the HSA and BSA concentrations. From the slopes of these linear relationships, the number of AOT ions per protein required for the maximal protection (NAOT) was approximately estimated to be 8.0
SDS Protects BSA Structure in Heat Denaturation
Figure 6. Dependence of [AOT]REQ on the HSA (O) and BSA (b) concentrations in the thermal denaturation at 65 °C.
Figure 7. Dependence of [SDS]REQ on the HSA (O) and BSA (b) concentrations in the thermal denaturation at 65 °C.
for HSA and 5.0 for BSA. These numbers are a little smaller than the numbers of stoichiometric binding sites of albumins against anionic surfactants.1-4 The concentrations of the intercepts in Figure 6 appear likely not to indicate an ineffective concentration or an unbound concentration of the surfactant in the binding process. This is because the protective effect, which is not so different from the maximal effect, evidently appears for the HSA structure even at 0.05 mM AOT, lower than the concentration of the intercept, as can be seen in Figure 5. A rather great protective effect appears in the coexistence of even quintuple AOT of HSA or BSA, probably by the binding of such a small amount of AOT. It should be noted that such a small amount of AOT ion, the molecular weight of which is about 1/150 of those of the proteins (66000-67000), functionally protects the protein structure. In the present study, the protective effect of SDS on the HSA helicity was also examined. The effect of SDS on the HSA helicity appeared in a manner similar to that on the BSA helicity,11 as a whole. A distinct difference appeared in the protein concentration dependences of [SDS]REQ. Figure 7 shows the dependences of [SDS]REQ on the HSA and BSA concentrations. The value of NSDS was 8.5 and 3.8 for HSA and BSA, respectively. The values of NSDS and NAOT for HSA were larger than those for BSA. This difference is discussed later. The concentrations of AOT and SDS, where the protective effect is observed, are too low not only to form their micelles but also to form aggregates on the protein polypeptides. Then the protein structures might be stabilized by a specific function, probably a cross-linking function of the bound AOT ions. The term of the crosslinking function was first proposed by Markus et al.7,8 This is the hypothesis that the native conformation is
Langmuir, Vol. 21, No. 12, 2005 5527
stabilized by a cross-linking function of anionic surfactant ion between a group of nonpolar residues and a positively charged residue located on different loops of the protein. The present results suggest that such linkages at particular sites can protect the protein structure in the thermal denaturation. The story of cross-linking may be too fabled for someone. However, the same orientation of dodecyl sulfate ion has been verified in the case of lysozyme by X-ray crystallography: the hydrophobic chain makes hydrophobic contacts with particular residues, while the hydrophilic group is interacting with charged residues.34 Each of the serum albumins has three domains, each consisting of a large double loop, a short connecting segment, a small double loop, a long connecting segment (hinge), another large double loop, and a connecting segment to the next domain, as first proposed by Brown.27 The helical structure is formed at most of these moieties.27 The total amount of helices in the large loops corresponds to the helicity of 45% as compared to the original helicity of 66%. The helices in the large loops have been considered to be stable, since the other helices in the C-terminal and in the connecting segments between the domains and between the loops are susceptible to various effects such as pepsin digestion.12,25,29,35-38 Also in the thermal denaturation up to 65 °C, the helices in the large loops are expected to remain intact.11 In the thermal denaturation, therefore, the protective effect on the helical structures might be expected to occur in the helical moieties, which are susceptible to various effects. As can be anticipated from Brown models of HSA and BSA,27 most of the helical rods are considered to be originally stabilized by making pairs. Some of these pairs have been verified by the recent X-ray crystallographic study of HSA.32,33 Most of the helical rods are amphiphilic.11 The hydrophilic parts of the amphiphilic rods contain Lys and/or Arg residues,11 which can electrostatically interact with anionic hydrophilic groups of anionic surfactants. This situation must be favorable for the above cross-linking of anionic surfactants. The protective phenomena have conspicuously been observed for the serum albumins.9-11 This might be due to the fact that these proteins are rich in pairs of amphiphilic helical rods.27 Indeed, the aforementioned helical moieties, which are susceptible to various effects, consist of pairs of helical rods.11,27 The cross-linking of the surfactant monomer might occur between a group of nonpolar residues and a positively charged residue located on several sets of amphiphilic helical rods in the proteins. Here, we see why the values of NSDS and NAOT for HSA were larger than those for BSA. It is already clear that only anionic surfactants show the protective phenomena for the structures of these serum albumins.11 In the case of the single-tailed anionic surfactants of alkyl sulfate, the surfactant with a longer chain is more effective for the protection and refolding of the protein structure.11 Needless to say, the present protective function is due to the amphiphilic nature of the anionic surfactant: each of the hydrophilic and hydrophobic groups plays an important role. It is indispensable in the protective effect that the hydrophilic group of the anionic surfactant interacts with a positively charged residue in the helical rod, as stated (34) Yonath, A.; Prodjarny, A.; Honig, B.; Sielecki, A.; Traub, W. Biochemistry 1977, 16, 1418-1424. (35) Braam, W. G. M.; Hilak, M. C.; Harmsen, B. J. M.; Van Os, G. A. J. Int. J. Pept. Protein Res. 1974, 6, 21-29. (36) Hilak, M. C.; Harmsen, B. J. M.; Braam, W. G. M.; Joordens, J. J. M.; Van Os, G. A. J. Int. J. Pept. Protein Res. 1974, 6, 95-101. (37) Feldhoff, R. C.; Peters, T., Jr. Biochemistry 1975, 14, 45084514. (38) Reed, R. G.; Feldhoff, R. C.; Peters, T., Jr. Biochemistry 1976, 15, 5394-5398.
5528
Langmuir, Vol. 21, No. 12, 2005
above. Some helical rods of BSA seem to miss positively charged residues such as Lys, which exist in the corresponding positions in HSA.10 Therefore, the protective effects of these anionic surfactants on the BSA structure are less than those on the HSA structure. Indeed, the helicity of HSA is completely recovered to the original value upon cooling to 25 °C after the heat treatment, but for BSA this is not the case, as mentioned above. Such a difference in the numbers of positively charged residues in the helical rods of these proteins might be related to the fact that the values of NSDS and NAOT for BSA are smaller than those for HSA. Although a surfactant is generally recognized to fulfill its function in the aggregated state such as a micelle, in the present Article we demonstrate the novel effect of a
Moriyama and Takeda
small amount of the double-tailed surfactant, the concentration of which is too low to form such aggregates. The surfactant ion must fulfill this function in a monomer state. This is a new and unique function presented by the amphiphilic nature of the ionic surfactant. We add that this function, which appears only in the presence of some denaturant or denaturing factor for a protein, has been unnoticed in many studies of surfactant-protein interaction carried out for half a century. Acknowledgment. We thank Chihiro Kobayashi, Mikio Ushiyama, Keiko Itano, and Yoshie Kawasaka for their helpful assistance in the experiment. LA050252J
